Movies-on-demand May
Significantly Change the Internet

(Theme issue of the Oct 1994 Bulletin of the American Society for Information Science)

Howard Besser
Visiting Asst. Professor
School of Information & Library Studies
University of Michigan


Current popular discourse views the "Information SuperHighway" as an enlarged and faster Internet. In this view increased content, primarily in the realm of entertainment, will create economies of scale that drive down costs and increase market penetration. Yet, the requirements of widespread online entertainment are likely to promote an infrastructure and culture that looks quite different than today's Internet.

In this short article we will use examples from the past to speculate about technological, economic, sociological, and cultural outcomes that are likely to flow from the establishment of online mass distributed entertainment. And we will show how this infrastructure and culture may differ significantly from that of today's Internet.

Technological Aspects

Popular sentiment holds that movies-on-demand will drive technological changes for the Information SuperHighway. The huge amount of storage space that video servers will require, the vast bandwidth needed to deliver videos to the home, and the scheduling and protocols needed for consumer-level access to this vast array of information will all pave the way for increased access to all types of information. But if we examine the optimal technological requirements for movies-on-demand, we will find that these probably are not optimal requirements for the type of information and query systems that currently reside on the Internet.

Delivery of digital video is extremely consumptive of bandwidth, even when highly compressed. Much of the discussion of "fiber to the home" has been prompted by the bandwidth needs of video. Because these bandwidth requirements dwarf those of all other types of information currently under discussion, a number of information deliverers have begun raising the issue of abandoning the practice of symmetrically balanced networks. From the distributor's point of view, a symmetrical configuration looks like a waste of bandwidth. The optimal network configuration for movies-on-demand would involve very high bandwidth into the home and very low bandwidth going back out (just enough to handle selection and billing functions).

The Internet, on the other hand, has grown up with symmetrical bandwidth as a given, and much of the Internet culture is based around this idea of symmetrical bandwidth. (For instance, the notion that anyone can easily become an information provider follows from the technological capability that all users have access to as much bandwidth going out as coming in.)

An asymmetrical design implies that only a limited number of sites will have the capability of outputting large volumes of bandwidth onto the Information SuperHighway. If such a configuration were to become prevalent, it would be likely to have several far-reaching results. It would inevitably lead to some form of gatekeeping. Managers of those sites will control all high-volume material that can be accessed. And for reasons of scarcity, politics, taste, or personal/corporate preference, they would make decisions on a regular basis as to what material would be accessible and what would not. This kind of model resembles broadcast or cable television much more so than it does today's Internet.

Issue of Mass -- Background and Historical Examples

Current experiments with movies-on-demand are primarily cooperative ventures between large companies in the cable, telephone, and entertainment fields (TCI, Viacom, Bell Atlantic, US West, AT&T, Time Warner, Disney). Eventually these companies expect to offer similar services throughout the country, and to realize their profits from mass distribution. They plan to stock popular titles and many plan to move extra copies of the most popular ones onto servers closest to the communities being served.

This is very different than the type of information we have come to expect on the Internet. Virtually all information planned for delivery as movies-on-demand is available through other distribution channels (tape rental, movie theaters, television), while most information currently on the Internet is only available from that single source. Copies of most movies-on-demand will reside on many different servers; content on the Internet today seldom is mounted on even a handful of sites.

Early cable television promoted narrowcasting (providing information to a relatively small select group) with local channels for government, schools, and public access. But as competition for channels began to rise, most of the narrowcasting that has survived relies upon the creation of a significant sized audience by combining many small local audiences. Thus cable narrowcast channels (devoted to Health, Travel, Seniors, etc.) still owe their survival to attracting a large audience (in the tens or hundreds of thousands instead of in the millions).

Narrowcasting is viable (perhaps even dominant) on the Internet because the system is decentralized and almost anywhere the information resides is equally accessible. But the introduction of "preferred servers" with much higher bandwidth delivery capabilities will create competition at those servers, and this competition will inevitably be won by those with either more economic clout or demonstrated mass appeal.[+]

Because fixed costs (such as digitization and storage) can be spread among more usage, distributors of movies-on-demand will realize their greatest profits on the most popular videos and will tend to favor them. This is similar to the "best-seller" phenomenon in bookstores, and can have consequences similar to those facing bookstores (as will be explained later).

Video-on-demand proponents promise "500 channels" and many claim that this will result in a far greater diversity of programs than currently available from any source. While, at least in theory, a vast selection of user choice is possible, diversity of choice is not an inevitable outcome of increased server and delivery capacity. We need only look at similar promises in the recent past to see that the diversity promised by a larger number of channels may prove to be a hollow promise.

The 1967 report of the President's Task Force on Communications Policy[6] made a series of recommendations on the role that should be served by emerging cable television systems. The industry should be structured "to cater to as wide a variety of tastes as possible, the tastes of small audiences and mass audiences, of cultural minorities and of cultural majorities. Television should serve as varied as possible an array of social functions, not only entertainment and advertising, ... but also information, education, business, culture, and political expression. ... Television should provide an effective means of local expression and local advertising, to preserve the values of localism, and to help build a sense of community... [P]olicy should guard against excessive concentration in the control of communications media."

A February 1973 report on the Future of Cable TV by the National Science Foundation was enthusiastic about what cable TV would offer. "Public access channels available to individual citizens and community groups. ... Churches, Boy Scouts, minority groups, high school classes, crusaders for causes -- can create and show their own programs. With public access, cable can become a medium for local action instead of a distributor of prepackaged mass-consumption programs to a passive audience. New services to individual subscribers, such as televised college courses and continuing education classes in the home. Cable's capability for two-way communication between viewer and studio may in time permit doctors to participate in clinical seminars at distant hospitals, or enable viewers to register their opinions on local issues. ... Public and private institutions might build their own two-way cable networks or lease channels to send x-rays among hospitals, exchange computer data, and hold televised conferences."[7] They listed key features that cable would offer including: "Instruction for homebound and institutionalized persons, Preschool education, High school and post-secondary degree courses in the home, Career education and in-service training, Community information programming, Community information centers, and Municipal closed-circuit applications."[8] The prediction made for cable television more than two decades ago sound remarkably like the predictions being made for the Information SuperHighway today. When listening to today's predictions, we should keep in mind how empty those proved for cable.

Independent video producers have welcomed each new distribution technology as a likely channel to get their programs out to the public. Yet each new distribution channel (cable TV, MTV, video stores) has proved disappointing, and eventually led to a solidifying of control by the largest producers. In the early 1980s, for example, video rental stores were touted as a great potential outlet for independent producers and distributors. Yet by the late 1980s it was clear that even in a region known as being especially receptive to independent media (the San Francisco Bay Area), the overwhelming majority of video rentals were large budget Hollywood productions. According to statistics kept by Peter Moore of the Captain Video chain, over 80% of the rentals at any given time came from the top 25 titles.[+]

Issue of Mass -- Effects

Because distributors will view their audience as a "mass" audience, a number of results are likely. First of all, information distributors will favor non-controversial programs (for fear of alienating part of their audience). In recent years we have seen the extreme version of this, where controversial programs have been eliminated from network television and radio, cable, local broadcast stations, and even art museums due to pressure from various organizations. But, perhaps less obvious is the fact that the overwhelming majority of programming focuses on elements that appeal to most people but don't offend anyone (the least-common-denominator), and this is due to the orientation towards a mass audience.

For similar reasons, programs designed for mass consumption will be favored over those perceived as having a relatively narrow appeal. The result is likely to be a lack of diversity and an emphasis on mass-appeal items. For an example of possible long-term results from this phenomenon we can examine what is currently happening in bookstores and video stores around the country. Independent book and video stores are rapidly being replaced by chain stores. Independents tend to offer a wide diversity of material, and in some ways use popular, mass-appeal items to subsidize more esoteric works. Chain stores, on the other hand, tend to carry little other than popular items. Because of the economies of scale realized by stressing mass-appeal items,[++] chains are putting the independents out of business, and it is getting more difficult to find items that don't have much mass appeal. If this carries over onto the Information SuperHighway, we can expect that what may start out as diverse offerings will, for economic reasons, soon turn into relatively homogenous programs with mass appeal.

Other portions of the "mass" phenomenon have parallels with book and video stores. Upholding a tradition they share with libraries, independent bookstores tend to take strong stands against censorship. Chain stores, on the other hand, tend to "not want to offend", and avoid carrying controversial items. (For example, after the death threat to Salman Rushdie, many independents carried both books and displays of Satanic Verses, while most chains refused to even carry it.)

Chain stores tend to deal almost exclusively with major publishers and distributors who can offer them better (volume) discounts and less paperwork. Independents tend to be one of the few venues for small presses or independent videos. A decade ago we saw legislative attempts to favor large studio productions over independent productions in proposals to "tax" blank tapes.[9] While these efforts were designed to compensate major producers for illegal copying, in effect they amounted to an attempted fund transfer from independent producers (who spent a more significant percentage of their budget on blank tape) to the large studios who were to receive the "tax" distribution based upon perceived market share.

Independents tend to be close to their communities, and cater to odd tastes within them. Chains, on the other hand, tend to focus on national tastes, and ignore items that may cater to primarily local or regional tastes. Some chains have been accused of trying to impose their perception of (bland) national tastes upon local communities. Blockbuster,[+++] for example, refuses to carry programs it deems controversial because of sexual or political themes, even in communities that do not find those themes offensive.[10] For a generation, this same phenomenon of "mass" allowed broadcast television to dominate the discourse over what constitutes national taste. This is likely to carry over onto the Information SuperHighway, with video-on-demand service providers imposing their national standards upon each local community they enter. Information providers will claim that this will be done for purely economic reasons -- it is not cost-justifiable to spend a fortune digitizing and mounting a program that will only be of interest to a few communities. But the results will be the same.


Though mass consumption entertainment-oriented content will speed the development of the Information SuperHighway, it is possible that many of the elements that current Internet users consider vital will disappear in the new infrastructure. Though the average consumer will have many more options than now available from their home television, attempts at mass distribution will likely favor mainstream big-budget programs over those that are controversial or appeal to a narrower audience. It is possible that diversity available from all sources will decrease and independent productions will be even further marginalized. And the adoption of an assynchronous architecture (a ten-lane highway coming into the library or home with a tiny path leading back out) would pose a significant barrier to those seeking to be information providers, and would favor a model of relatively passive consumption.


Clifford Lynch first alerted the author to the idea of asymmetrical bandwidth. Discussions with Lester Weiss and Peter Moore of the Pacific Film Archive helped develop a number of the ideas raised here. Eva Garcelon provided research assistance. The views put forward in this article are solely those of the author.

[+] Steve Cisler has pointed out that videos that appeal to very few people in any one community might, on a national level, have a large enough audience to sustain a niche server (such as a "Cult Film Server"). But, until the cost of video bandwidth becomes negligible, the cost of delivering these videos across country will make them appear prohibitively expensive to the consumer (at least in comparison with mass market videos).

[6]United States. President's Task Force on Communications Policy. Final report, August 14, 1967. [Washington]: GPO, 1968, [chapter VII, pages 3-4]. (commonly known as the Rostow Report)

[7]Baer, Walter S. Cable Television: A summary overview for local decisionmaking, National Science Foundation Research Applied to National Needs Program, 134-NSF, Santa Monica: Rand, February 1973, page 2.

[8]ibid, page 6.

[+] Peter Moore, currently of Pacific Film Archive (Berkeley, CA) in data and statistics shared with the author.

[++] This is similar to the phenomena of "cherry picking" that in recent years has put intense pressure on the Regional Bell Operating Companies (RBOCs). Companies offering phone connections to business customers in metropolitan areas have "skimmed off" the most lucrative customers by offering cheaper prices (and can do so because they are not using these customers to subsidize other services). By removing the RBOCs' highest-paying customers, this has forced regulatory agencies to consider changing the fee structures for all customers.

[9]see, for example, Tax on home videotaping is urged, New York Times v131 (Thu, April 22, 1982):24(N), C21(L), col 2, Holsendolph, Ernest, Legislative Plan to Tax Video Recording Gear, New York Times v131 (Fri, March 12, 1982):A30, and Raspberry, William. No to the Betamax tax (column), Washington Post v106 (Fri, April 29, 1983):A29, col 1, and Lardner, James, Fast Forward: Hollywood, the Japanese, and the onslaught of the VCR, New York: Norton, 1987.

[+++] Which the NY Times has called "the largest and most profitable movie rental chain in the United States" (Portfolio: Huizenga's Major Holdings, NY Times, April 4, 1994, page B12) and the Wall Street Journal calls "the MacDonald's of US video rentals" (Brannigan, Martha. Blockbuster plans a drive to become the MacDonald's of U.S. video rentals, Wall Street Journal (Fri, June 10, 1988):28(W), 29(E), col 3).

[10]see, for example, Blockbuster to shun video of 'The Last Temptation,' Wall Street Journal (Fri, June 23, 1989):B2(W), B4(E), col 6, and Blockbuster, San Francisco Weekly (March 29, 1991), page 15. Last modified: 3/11/1997